LLMO — Large Language Model Optimization
Working on the corpus a language model has already trained on — mentions, reviews, listings, partner sites, press — so the model has the right associations baked in by the time a user asks about your category.
LLMO is the upstream half of the AI presence stack. Models don't just retrieve at query time; they carry priors from training data. LLMO shapes those priors. Practically: identifying which sources a model already weights heavily in your category, then methodically becoming part of those sources — through digital PR, listings, partnerships, reviews, expert citations and signed thought leadership.
Retrieval-augmented generation can fix a few things at query time, but the model's baseline associations are set during training. LLMO is how you change those baselines.
A B2B vendor identifies G2, Gartner peer reviews and three industry publications as the sources Claude and GPT-4 weight in their category. Six months of focused PR turns those sources into reliable citations of the vendor — and the vendor starts appearing in 'top 5 vendors for X' AI replies.
GEO
The discipline of optimising a brand's content, structure and external footprint so generative AI engines (ChatGPT, Gemini, Perplexity, Copilot, Google AI Mode) discover, understand and cite it inside their answers.
AEO
Optimising content to be the source an answer engine — Google AI Overviews, Bing Copilot, Perplexity, ChatGPT search — quotes and links to in its synthesised reply.
Citation Rate
The percentage of AI answers in a defined prompt set where your brand is named or linked to as a source — the cleanest single metric for AI visibility.
Do you know what AI
says about you?
Request an audit and discover how your brand appears when customers, partners and investors ask AI for solutions, recommendations, comparisons or vendors in your category.
- 01Analysis across ChatGPT, Gemini, Perplexity, Copilot and Google AI Mode
- 02Real comparison with your main competitors
- 03Citations, mentions and source review
- 04Detection of errors and incomplete information
- 05Content and authority opportunities
- 06Executive 30 / 60 / 90 day roadmap